Climă/Teme de date INSPIRE (în conformitate cu anexele la Directiva 2007/2/CE ) / 2023_Romania_GHG Projections.pdf

𝑦 −𝑦0 for 𝑘 = 𝑇 → 𝑦̂𝑇 = 𝑦1 + (𝑇 − 1) ⋅ 𝛥 = 𝑦1 + (𝑇 − 1) ⋅ 𝑇 = 𝑦𝑇 (4) 𝑇−1 For extrapolation, the k series is continued until the last unit of time related to the researched horizon - the year of 2050. The evolution of the surfaces related to the specified areas has a linear projection over time that will be estimated by the method of absolute mean change. The average index method is applied if the surfaces have a decreasing trend and the projection by the average increase method generates negative terms or equal to zero. The adjusted values are determined with the following recurrence ratio: 𝑘−1 𝑦̂𝑘 = 𝑦1 ⋅ 𝐼 , 𝑘 = 1, 𝑇 (5) where 𝑦1 is the term considered as the basis for adjustment (year 1989 in the reference scenario-WOM (business as usual), year 2005 in the case scenario with existend measures-WEM and year 2012 in the case of the scenario with additional measures-WAM) and represents the average index calculated as simple geometric mean of relative changes based on in the chain: 𝑇−1 𝑦2 𝑦3 𝑦4 𝑦𝑇−1 𝑦 𝑇−1 𝑦𝑇 𝐼= √𝑦 ⋅ 𝑦 ⋅ 𝑦 . . . . . 𝑦 ⋅𝑦 𝑇 = √𝑦 (6) 1 2 5 𝑇−2 𝑇−1 1 The first and last adjusted value are equal to the first and last empirical value, as follows: 1−1 for 𝑘 = 1 → 𝑦̂1 = 𝑦1 𝐼 = 𝑦1 ; (7) 𝑇−1 𝑇−1 𝑇−1 𝑦 for 𝑘 = 𝑇 → 𝑦̂𝑇 = 𝑦1 ⋅ 𝐼 = 𝑦1 ⋅ ( √ 𝑇 ) = 𝑦𝑇 . (8) 𝑦 1 Based on the index, the average rate (R) with the relation is calculated: 𝑅(%) = 𝐼(%) − 100 (9) and together with the average level and the absolute average change, it contributes to the statistical characterization of the time series. Once the surfaces have been estimated, we start from the hypothesis that the 17 quantitative level of GHG depends on the surface of the specified area and/or on the time variable, a function that is determined using the regression method. The regression method is a statistical approach for determining the connection between variables using mathematical functions, called regression functions. A mathematical expression obtained from the processing of experimental data that approximates the interdependence between two or more variables of a system or process is referred to as a regression function. When the connections between the various variables cannot be demonstrated theoretically accurately enough, a regression function must be determined. The following processes were involved in estimating a mathematical model that was linked to a certain type of projection: (i) intuition of mathematical relations of the models based on the graphical representation of the correlation; (ii) application of the Pearson coefficient to determine the intensity and direction of the link between GHG emissions-removals as a function of time and/or specific areas of different land categories. The Pearson coefficient ranges from -1 to +1; the closer it is to zero, the lower the bond's intensity, and the closer it is to one, the stronger the relationship. The relationship is directly proportional if the values are greater than zero, and inversely proportional if the values are negative; (iii) estimating the parameters for each model, most often using the least squares method; (iv) choosing the model that most accurately represents the relationship highlighted by the data (minimum criterion); (v) testing the significance of the chosen model and the coefficients of the functions found; (vi) assessing the model's significance and the coefficients of the functions identified; (vii) economic interpretation of the tested parameters; (viii) use of the model for simulations and projections. There are one-factor regression (single) and multi-factor regression (multiple), depending on the number of influencing factors for the resulting characteristics. Unifactorial regression considers quantifying dependencies in the form of unifactorial models, testing hypotheses, and demonstrating predictive calculations. The unifactorial model describes the relationship between the two variables, x and y, the other factors being considered with constant action. The theoretical equation of this regression is of the form: 𝑦𝑖 = 𝑓(𝑥𝑖 ) + 𝑢𝑖 , 𝑖 = 1, 𝑛 (10) where 𝑢𝑖 it represents the action of other factors (disturbance), and 𝑛 the number of observations. The estimation of the parameters of this model is done using the least squares method, which involves minimizing the sum of the squares of the deviations of the empirical values (𝑦𝑖 ) from the estimated values (𝑦̂𝑖 ): ∑𝑛𝑖=1(𝑦𝑖 − 𝑦̂𝑖 )2 → 0 (11) In unifactorial linear models, the resultant variable depends on a single factor by the relation: 𝑦𝑖 = 𝑎 + 𝑏 ⋅ 𝑥𝑖 + 𝑢𝑖 ,𝑖 = 1, 𝑛 (12) The estimation of the parameters of this model can also be done using the least squares method, which involves minimizing the sum of the squares deviations of the empirical values (𝑦𝑖 ) from the estimated values (𝑦̂𝑖 ): 2 ∑𝑛𝑖=1(𝑦𝑖 − 𝑦̂𝑖 )2 = ∑𝑛𝑖=1(𝑦𝑖 − 𝑎̂ − 𝑏̂ ⋅ 𝑥𝑖 ) → 0 (13) where 𝑎̂ and 𝑏̂ are the estimators of the linear model parameters. Minimizing the amount involves determining the stationary points that result from solving the system obtained by canceling the partial derivatives of the function: 2 𝐹(𝑎̂, 𝑏̂) = ∑𝑛𝑖=1(𝑦𝑖 − 𝑎̂ − 𝑏̂ ⋅ 𝑥𝑖 ) (14) we arrive at the following system of equations: 18 𝑛𝑎̂ + 𝑏̂ ∑ 𝑥𝑖 = ∑ 𝑦𝑖 { (15) 𝑎̂ ∑ 𝑥 + 𝑏̂ ∑ 𝑥𝑖2 = ∑ 𝑥𝑖 𝑦𝑖 𝑖 and after performing the calculations the following estimation relations are obtained: ∑𝑛 𝑛 2 𝑛 𝑛 𝑖=1 𝑦𝑖 ∑𝑖=1 𝑥𝑖 −∑𝑖=1 𝑥𝑖 ∑𝑖=1 𝑥𝑖 𝑦𝑖 𝑛∑ 𝑛 𝑥 𝑦 −∑ 𝑛 𝑥 ∑ 𝑦𝑖 𝑛 𝑎̂ = 2 (16) 𝑏̂ = 𝑖=1𝑛 𝑖 𝑖 2 𝑖=1𝑛 𝑖 𝑖=1 2 (17) 𝑛 ∑𝑛 2 𝑛 𝑖=1 𝑥𝑖 −(∑𝑖=1 𝑥𝑖 ) 𝑛 ∑𝑖=1 𝑥𝑖 −(∑𝑖=1 𝑥𝑖 ) In the case of multifactorial regression, a resultant variable is defined according to several factorial variables which implies the use of multifactorial regression: 𝑦 = 𝑓(𝑥1 , 𝑥2, . . . , 𝑥𝑛 ) + 𝑢𝑖 , 𝑖 = 1, 𝑛 (18) where 𝑢𝑖 it represents the action of other factors (disturbance), and 𝑛 the number of observations. If the connection between each factor and the resultant variable is linear, then the regression function becomes: 𝑦𝑥1 ,𝑥2, ...,𝑥𝑛 = 𝑎0 + 𝑎1 ⋅ 𝑥1 + 𝑎2 ⋅ 𝑥2 +. . . +𝑎𝑛 ⋅ 𝑥𝑛 (19) in which 𝑎0 is the parameter that concentrates the influence of the unregistered factors, considered with constant action; 𝑥1 , 𝑥2, . . . , 𝑥𝑛 are the factorial variables included in the research report; and 𝑎1 , 𝑎2, . . . , 𝑎𝑛 are regression coefficients, which show how much the resultant variable changes when the factorial variable changes by one unit. The estimation of the parameters of this model can be done using the least squares method, which involves minimizing the sum of the squares of the deviations of the empirical values (𝑦𝑖 ) from the estimated values (𝑦̂𝑖 ): ∑𝑖(𝑦𝑖 − 𝑦̂𝑖 )2 = ∑𝑖(𝑦𝑖 − 𝑎̂0 − 𝑎̂1 ⋅ 𝑥1 − 𝑎̂2 ⋅ 𝑥2 −. . . −𝑎̂𝑛 ⋅ 𝑥𝑛 )2 (20) where, (𝑎̂𝑖 )𝑖=0,𝑛 are the estimators of the linear model parameters. Minimizing the amount involves determining the stationary points that result from solving the system obtained by canceling the partial derivatives of the function: 𝐹(𝑎̂0 , 𝑎̂1 , 𝑎̂2 , . . . , 𝑎̂𝑛 ) ∑𝑖(𝑦𝑖 − 𝑎̂0 − 𝑎̂1 ⋅ 𝑥1 − 𝑎̂2 ⋅ 𝑥2 −. . . −𝑎̂𝑛 ⋅ 𝑥𝑛 )2 → 0 (21) The following system of equations is constructed based on empirical data: 𝑛 ⋅ 𝑎0 + 𝑎1 ⋅ ∑𝑖 𝑥𝑖1 + 𝑎2 ⋅ ∑𝑖 𝑥𝑖2 +. . . +𝑎𝑛 ⋅ ∑𝑖 𝑥𝑖𝑛 = ∑𝑖 𝑦𝑖 2 𝑎0 ⋅ ∑𝑖 𝑥𝑖1 + 𝑎1 ⋅ ∑𝑖 𝑥𝑖1 + 𝑎2 ⋅ ∑𝑖 𝑥𝑖1 ⋅ 𝑥𝑖2 +. . . +𝑎𝑛 ⋅ ∑𝑖 𝑥𝑖1 ⋅ 𝑥𝑖𝑛 = ∑𝑖 𝑥𝑖1 ⋅ 𝑦𝑖 (22) ....................................................................................... 2 2 { 𝑎0 ⋅ ∑𝑖 𝑥𝑖𝑛 + 𝑎1 ⋅ ∑𝑖 𝑥𝑖1 + 𝑎2 ⋅ ∑𝑖 𝑥𝑖𝑛 ⋅ 𝑥𝑖2 +. . . +𝑎𝑛 ⋅ ∑𝑖 𝑥𝑖𝑛 = ∑𝑖 𝑥𝑖𝑛 ⋅ 𝑦𝑖 Based on the theoretical elements presented, it was necessary to decide on the statistical significance of the regression parameters and the validity of the adopted model. If the regression coefficients of the estimated models belong to a confidence interval that also contain the value of 0, the model was abandoned. After verifying the significance of the parameters, the next step was to validate the statistically specified model using the ANOVA test. Validating the model, which was assumed to be significant for the regression parameters, was done by interpreting a specific p-value test. Once validated, the model was used in projection simulation. 19 1.5 Methodology for GHG projection estimations in the Waste sector Waste disposal in landfills has a direct impact on the environment, including by generation of GHG emissions. Waste degradation is a complex of chemical and biological reactions, whose result is the generation of biogas with the basic composition: CH4 and CO2. The biogas decomposition and removal continue over 10 to 30 years. A 50% of degradable organic waste is decomposed during the last 10 years, 12.5% of the remaining decompose over 30 years. Projections of GHG emissions are made in accordance with the IPCC 2006 Guidelines. To achieve emission projections, we started from the basic parameters on which base emissions from the waste sector are estimated in the National Greenhouse Gases Inventory. Methodology for GHG emission projections for wastewater sector CH4 emissions from domestic and commercial wastewater treatment and sludge The IPCC 2006 documentation applied to Romania's conditions highlights the existence of the following possible sources of methane emissions: • Household wastewater collected in sewerage networks and discharged into a collector without treatment. These waters undergo an aerobic self-purification process with minor methane emissions • Household wastewater collected in sewerage networks and treated in city treatment plants using aerobic and anaerobic biological processes • Household wastewater from non-sewer areas is collected in septic tanks or cesspools, where aerobic and anaerobic processes are carried out, and that cannot be controlled. The sources of methane as result of the anaerobic processes are more developed in high temperature areas and when the interval between vacuums is high. The large wastewater treatment plants are equipped with active sludge basins containing aerobic, anoxic and oxic areas. The oxic areas can become anoxic and even anaerobic if the treatment process is not properly driven. In this case, methane emissions may occur. The sludge from the treatment plants is subjected to stabilization processes that can be anaerobic or aerobic. Biogas with a high methane content is produced in the case of anaerobic treatment of sludge. Typically, biogas is used in electricity or heat generation facilities. In case of malfunctions, methane can be released in the atmosphere. The above-mentioned household wastewater can be a significant source of methane emissions that are coming from: • Population not connected to sewerage system • Population connected to sewerage system, with treatment. The CH4 emissions from the treatment of domestic wastewater depend directly on the population, biochemical oxygen consumption, methane emission factor (kg CH4/kg BOD), the using of wastewater treatment plants, the organic component removed by sludge and the amount of CH4 recovered from wastewater. 20 CH4 Emissions from Industrial Wastewater Treatment The estimation of methane production potential from industrial wastewater treatment was based on the concentration of organic components in wastewater, the volume of wastewater and the treatment of industrial wastewater in anaerobic systems. The sectors with the highest potential methane emissions from wastewater treatment are Pulp and paper manufacturing, Food (dairy products, beer, alcohol) and Crude oil refining. The CH4 emissions from industrial wastewater treatment depends directly on specific industrial production, the amount of wastewater generated, chemical oxygen consumption, methane emission factor (CH4/kg COD kg), the organic component removed by sludge and the amount of CH4 recovered. N2O emissions from wastewater treatment N2O emissions from human waste are related to the presence of nitrogen from proteins consumed. According to IPCC 2006, N2O emissions are: • Direct emissions in the case of modern wastewater treatment plants equipped with nitrification and denitrification technology • Indirect wastewater emissions after their discharge into the receptors. Considering the small share of modern wastewater treatment plants, only indirect N2O emissions are reported in NGHGI. Indirect N2O emissions depends directly on population, annual protein consumption per capita and on protein factors and emission factors for N2O emissions.  21 2. ASSUMPTIONS CONSIDERED FOR GHG EMISSIONS PROJECTIONS 2.1 Macroeconomic indicators To define the assumptions regarding the evolution of Romania in the period 2021÷2050, a SWOT analysis was made for the period 1989÷2020 targeting: • economic growth • demographic development • social development • structural adjustment of the economy • structural adjustment of the industry • technological modernisation and reduction of energy intensity in industry, agriculture, construction • transport sector development and modernisation • services sector development and modernisation • development and modernization of living conditions. In order to estimate the forecast for 2021÷2050 period it should be taken into account the average annual GDP growth rate indicated by the National Commission for Strategy and Prognosis in the Report Projection of the main macroeconomic indicators (from October 2022) for period 2019÷2030 and years 2035 (1.71%), 2040 (1.17%) and by the European Commission in the greenhouse gas projections made for the year 2023 (table 2-1). Table 2-1 Average annual GDP growth rate (%), 2005÷2050 Year 2005 2018 2019 2020 2021- 2026- 2031- 2036- 2041 - 2046 - 2025 2030 2035 2040 2045 2050 Inventory Projections National Commission for Strategy 4.7 4. 4 4.1 -3.75 5.05 3.52 2.395 1.86 1.56 1.4 and Prognosis (October 2022) Parameters recommended by the - - - -3.75 3.63 3.04 2.16 1.27 - - European Commission (year 2022) Table 2-2 presents the GDP growth rates for 2020÷2050 period provided by the National Commission for Strategy and Prognosis, which were considered in the estimation of GHG projections. Table 2-2 Average annual GDP growth rate (%) considered in the GHG projections, 2005÷2050 Year 2005 2018 2019 2020 2021- 2026- 2031- 2036- 2041 - 2046 - 2025 2030 2035 2040 2045 2050 Historic values Projections Growth rate, % 4.7 4.4 4.1 -3.75 5.05 3.52 2.395 1.86 1.56 1.4 The positive evolution of the Romanian economy in the period 2000÷2018 led to an increase by 2.32 % in the gross domestic product per capita (9,814 Euro2016/capita at 2018 level). In the period 2007÷2016 the gross domestic product per capita increased by almost 20% compared to the European Union average. So, if in 2007 it stood at 40% of the average, in 2016 t